This book focuses on the adaptation of speech recognizers to noisy or reverberant environment. Therefore, three corpora in different noise and reverberation levels are presented.§Speech recognition is described in detail starting from the very basics. From feature extraction with Mel Frequency Cepstrum Coefficients (MFCC) and several variants of the§TempoRAl Patterns (TRAPs) all relevant details of speech recognition are explained.§In order to improve speech recognition even further all state-of-the-art speech recognizer adaptation techniques are featured: Methods like maximum a posteriori (MAP), maximum likelihood linear regression (MLLR), and constrained MLLR (CMLLR) are described in detail. Moreover, the Baum-Welch algorithm is used to interpolate the transition probabilities of the hidden Markov models of the speech recognizer.§By application of the adaptation techniques and artificially reverberated data a significant improvement of the recognition rate is achieved.§This book is an ideal introduction into speech recognition and speech recognizer adaptation techniques.
Maier
Speech Recognizer Adaptation jetzt bestellen!